PROJECT SUMMARY We and others have shown that a subset of tumor cells capable of regenerating new tumors, termed variously as tumor-initiating cells (TIC) or ?cancer stem cells? (CSC), are comparatively resistant to current systemic chemo- and radio-therapies relative to the bulk of the tumor. As such, the guiding premise for our work over the past decade, is that therapeutic targeting of processes required for survival or function of TIC will allow elimination of such cells at both the primary and metastatic sites, thereby enhancing response to current systemic therapies. However, it has become clear from several laboratories using current cell surface markers, enzymatic activity markers, and lentiviral signaling reporters for Wnt, Hedgehog, and STAT3-mediated signaling, that TIC are heterogeneous not only across tumors, but also show different behaviors, treatment responses, and reporter gene expression at the primary and metastatic sites within a given tumor model. Thus, it is critical to understand the diversity of TIC types both within and across tumors, as well as to understand how such cells differ in function between the primary and metastatic site, in order to target them effectively ? particularly in ?triple-negative? breast cancer (TNBC), for which there are no approved targeted therapies. We hypothesize that there exist molecularly distinct classes of TIC in TNBC. As a consequence, differential responses to individual chemotherapies depend, in part, on the class(es) of TIC present in a given tumor. If true, we further assert that identifying and targeting class-specific TIC functions may help overcome chemotherapy resistance. In order to identify candidate regulatory genes or pathways functioning in a spectrum of PDX-derived TIC isolated from primary and metastatic sites, we propose to determine first whether TIC derived from either the primary or metastatic site show identical, or different, patterns of activation of signaling networks implicated in TIC function using lentiviral fluorescent reporters specific for signaling pathways and transcription factors implicated in TIC function. We will then identify candidate molecular mechanisms required for TIC activity at the population and single cell levels using RNAseq. In order to determine whether in vivo targeting of TIC-related genes/pathways can augment response to carboplatin, and lead to elimination of different classes of TIC at both the primary and metastatic sites, we will evaluate the ability of FDA-approved agents targeting TIC to augment treatment response alone or in combination with carboplatin at the primary and metastatic sites, and assay the effect of treatment on TIC function in vivo. Results may be used to guide planning and execution of human clinical trials and inform precision medicine efforts.